Subrahmanya S

Automation Engineer with 3 years of experience at Siemens, specializing in end-to-end automation and high-quality software delivery using C#, Python and JavaScript. Proven track record of integrating advanced networking concepts and industrial device communication into scalable technical solutions. A proactive innovator and GenAI hackathon contributor, expert at leveraging GitHub Copilot and AI-driven workflows to accelerate development cycles. Combines deep technical curiosity with a "hardware-aware" engineering mindset to solve complex challenges in fast-paced, mission-critical environments.

AI-Driven DOM Analysis & Macro Synthesis Engine

• Leveraged Generative AI models to analyze HTML DOM structures, enabling the autonomous generation of reusable test components and page objects.
• Leveraged C# and Semantic Kernel to provide a seamless interface between generative AI models and enterprise-level automation frameworks.
• Increased automation coverage and reduced manual scripting efforts by approximately 70% through intelligent element mapping and script synthesis.

High-Concurrency JSON Data Transformation Tool

• Developed a robust C# WinForms utility designed for rapid, large-scale schema migration and multi-file synchronization in mission-critical environments.
• Automated the configuration lifecycle by engineering a batch-injection system that handles concurrent updates to hundreds of files with zero data loss.
• Drastically accelerated deployment workflows by replacing manual configuration management with a programmatic, error-resistant transformation pipeline.

Real-Time Visual-Linguistic Synthesis Pipeline

• Architected a high-throughput computer vision pipeline in Python to generate natural language descriptions from live-streamed visual data.
• Engineered a hybrid CNN-LSTM neural network utilizing a Dense-layer image encoder and a sequential language processor, integrated via a Concatenate layer to map high-dimensional visual features to linguistic patterns.
• Optimized temporal modeling using LSTM cells and the RMSprop optimizer, achieving high-accuracy captioning and low-latency performance in dynamic, real-world environments.

Vodafone - Automated Computer Vision Attendance Pipeline

• Designed and deployed a real-time biometric attendance system using Python and OpenCV to facilitate seamless, touchless time-tracking.
• Built a custom session-tracking engine that maps individual user "in-frame" duration to a centralized database for longitudinal data analysis.
• Integrated automated data visualization workflows to generate trend graphs and presence analytics, streamlining the administrative reporting process.

• English
• Hindi
• Kannada
• Telugu